Removal of fluoride from drinking water by a membrane coagulation reactor (MCR)

This paper studied the removal of fluoride from drinking water by a membrane coagulation reactor, and aluminum sulfate was used as the major chemical. In a jar test, the optimum pH value for removal of fluoride was found in the range from 6.0 to 6.7. In the laboratory-scale test, when proper dosages of aluminum sulfate and sodium hydroxide were added to the reactor, the concentrations of fluoride were reduced from 4.0 mg/L in raw water to less than 1.0 mg/L in finished water, and the concentrations of residual aluminum in finished water was no more than 0.05 mg/L. The dominant particle sizes of aluminum hydroxide in the mixed liquor were larger than 1 µm. The small particles whose sizes ranged from 0.2 µm to 0.4 µm were less, but they played an important role in the membrane fouling. The volume ratio of treated water to discharged sludge was 350, but it increased to 744 after the sludge was concentrated and dried at normal temperature for 41 days.

In the jar test, the optimum pH ranged from 6.0 to 6.7 when Al2(SO4)3 was used as the major chemical. In the laboratory-scale test, when a proper dosage of chemicals was injected into the MCR, the concentrations of fluoride could be reduced from 4 mg/L in raw water to below 1 mg/L in finished water and the concentrations of residual aluminum were less than 0.05 mg/L. Approximately 50 g of Al2(SO4)3 were consumed for removal of 1 g of fluoride. Although the dominant particles in the mixed liquor were Al(OH)3 whose sizes were larger than 1 µm and the volume percentage was as high as 94%, they played a small role in membrane fouling before SS in the mixed liquor increased to a critical value. However, the volume percentage of the smaller particles ranging from 0.2 µm to 0.4 µm was less than 3%, yet it played a significant role in membrane fouling. There was a critical SS during the operation period. The critical SS reached 11.9 g/L in the intermittent runs while it was only 5.4 g/L in the continuous runs. Intermittent runs were suggested for the practical application. The CF was about 350 in the intermittent runs when sludge was discharged from the reactor, but it increased to 744 after sludge was concentrated and dried at normal temperature for 41 days. Whether the dry sludge could be reused needs to be studied further.